Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-04T19:46:51.604848Z
Paper Citation Record · LEDGER
As of 10 August 2026, this Paper Citation Record lists 76 of 76 outbound references and 1 inbound Pith citation observation for arXiv:2509.09112.
A citation records a reference. It does not transfer a finding from one paper to another.
Typed states for the displayed outbound observations.
Source: paper_references, paper_reference_links, observed 2026-08-04T19:46:51.604848Z
One-hop event checks from named stored sources.
Source: scholarly_work_events, retraction_status_cache, observed 2026-08-09T06:31:02.800959+00:00
Pith citing papers itemized under the disclosed page cap.
Source: paper_references, paper_reference_links, observed 2026-08-06T04:16:06.800914Z
A source-named dated measurement, never combined with another source.
Source: pith, observed 2026-08-06T04:16:07.072616Z
76 of 76 outbound references displayed
External citation measurements
No source-named external measurement is stored.
Observation af1c3dac-5a6b-47f0-b9da-6114394187e5 · outbound
Character-Level Perturbations Disrupt LLM Watermarks OPT: Open Pre-trained Transformer Language Models
Reference 1
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation f95d0745-38c1-45f7-bf9d-1747d668a072 · outbound
Character-Level Perturbations Disrupt LLM Watermarks LLaMA: Open and Efficient Foundation Language Models
Reference 2
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 9e718246-4dfd-400e-9ab1-817f5907d1cc · outbound
Character-Level Perturbations Disrupt LLM Watermarks Openai chatgpt blog,
Reference 3
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation aeba42fb-1191-40ac-8275-8b3062c3f54a · outbound
Character-Level Perturbations Disrupt LLM Watermarks Deepseek terms of use,
Reference 4
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 73bbabcb-5d4a-413d-9124-68b9f30a9271 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Gemini api additional terms of service,
Reference 5
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 0dc0cfca-559d-4ebe-b84c-c1bbeb95ca34 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Grammarly
Reference 6
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 86c47f59-411a-45c1-b8f9-03162a0c3751 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Chatpaper
Reference 7
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 25a8acb9-e072-476f-a2ce-7293e0584fa2 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Transforming the future of music creation
Reference 8
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 23e1971f-30a2-44ef-896a-8681f59f4373 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Quickstart: Generate and verify an image’s watermark using imagen text-to-image (console)
Reference 9
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 610cf1b5-f003-4000-9641-276426b4deb3 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Self-supervised adversarial example detection by disentangled representation,
Reference 10
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3b04871c-e940-44a5-8b3b-f7d8352b813c · outbound
Character-Level Perturbations Disrupt LLM Watermarks Evaluating membership inference through adversarial robustness,
Reference 11
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 64306f4c-6f26-478c-a3af-4a189294a195 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Not all edges are equally robust: Evaluating the robustness of ranking-based federated learning,
Reference 12
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ca756dd5-3883-4cd9-92c7-7c9bd68bbd06 · outbound
Character-Level Perturbations Disrupt LLM Watermarks When better features mean greater risks: The performance- privacy trade-off in contrastive learning,
Reference 13
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1f4cd71c-4cb2-4b1b-a515-e188907bf284 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Dynamic graph unlearning: a general and efficient post-processing method via gradi- ent transformation,
Reference 14
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b946c5b4-04e9-4458-ba16-ee32c629e76b · outbound
Character-Level Perturbations Disrupt LLM Watermarks Unraveling privacy risks of individual fairness in graph neural networks,
Reference 15
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 4d14f793-8ff2-45e7-b6d5-339e9e2d1425 · outbound
Character-Level Perturbations Disrupt LLM Watermarks On the risk of misinformation pollution with large language models,
Reference 16
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation cd411e46-a3fb-49d2-8a12-4bdd5d8ab87c · outbound
Character-Level Perturbations Disrupt LLM Watermarks Spear Phishing With Large Language Models
Reference 17
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation dd80004e-74c3-433e-bef1-7e8fcf9fe996 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Chatgpt for good? on opportunities and challenges of large language models for education,
Reference 18
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6ef2c037-1a1a-42fe-9191-1cfbe991e1b9 · outbound
Character-Level Perturbations Disrupt LLM Watermarks A watermark for large language models,
Reference 19
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1f471d1d-fafd-4e8d-98d9-9a1f59772796 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Provable robust watermarking for AI-generated text,
Reference 20
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7fba4bf8-80ec-4bab-b539-049ecb010e5d · outbound
Character-Level Perturbations Disrupt LLM Watermarks Unbiased watermark for large language models,
Reference 21
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 681c2d0a-2566-4b9d-b959-ba06eb95b302 · outbound
Character-Level Perturbations Disrupt LLM Watermarks A resilient and accessible distribution-preserving watermark for large language models,
Reference 22
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 074e755d-8d82-4cfc-9609-5b9462d03de3 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Scalable watermarking for identifying large language model outputs,
Reference 23
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 219bb5de-11c5-4d2e-a0ab-b65e2e9f7158 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Robust distortion-free watermarks for language models,
Reference 24
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5a130255-fa94-4c60-aada-778758f4a513 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Undetectable watermarks for language models,
Reference 25
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 711f4715-12be-4419-ae98-4a1b7e30a767 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Bimark: Unbiased multilayer watermarking for large language models,
Reference 26
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d96a3aa7-4cf8-4296-b2e5-c40baf2907a1 · outbound
Character-Level Perturbations Disrupt LLM Watermarks A Certified Robust Watermark For Large Language Models
Reference 27
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 375488aa-3132-453d-8623-a8c37b759485 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Bypassing LLM watermarks with color- aware substitutions,
Reference 28
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6a8047f6-5637-484d-860f-005adf515442 · outbound
Character-Level Perturbations Disrupt LLM Watermarks b 4: A black-box scrubbing attack on LLM watermarks,
Reference 29
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2e6c8b32-61c5-48a6-86b2-fb5f686054e0 · outbound
Character-Level Perturbations Disrupt LLM Watermarks De-mark: Watermark removal in large language models,
Reference 30
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 01dcd123-91e0-498c-b77f-1eee60e2b138 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Watermark Smoothing Attacks against Language Models
Reference 31
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation b47e50ba-84ee-4318-81ef-6f983f7a5180 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Paraphrasing evades detectors of ai-generated text, but retrieval is an effective defense,
Reference 32
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 170903a3-188e-463e-b5f2-3ed59e5dcd1e · outbound
Character-Level Perturbations Disrupt LLM Watermarks On the reliability of watermarks for large language models,
Reference 33
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2bac9422-7bc1-44a2-a67e-2918d54d90a1 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Markmywords: Analyzing and evaluating language model watermarks,
Reference 34
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d78e126a-9892-4403-baef-81335802e8ef · outbound
Character-Level Perturbations Disrupt LLM Watermarks Textbugger: Generating adversarial text against real-world applications,
Reference 35
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ab842ec1-e927-44b5-b1b1-6832a794a607 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Textattack: A framework for adversarial attacks, data augmentation, and adversarial training in nlp,
Reference 36
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6656c59e-9302-4c0c-a6cb-e3934b8040d7 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Black-box generation of adversarial text sequences to evade deep learning classifiers,
Reference 37
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d3bcf567-a4fc-46af-be4b-6aa044c64693 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Exploring Gradient-Guided Masked Language Model to Detect Textual Adversarial Attacks
Reference 38
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 95f0f9bc-0bf2-42c1-beab-257e035e8796 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Masked language model based textual adversarial example detection,
Reference 39
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 97c9dea4-216b-488f-9676-6b630a7290d5 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Watermark stealing in large language models,
Reference 40
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d655e9cd-ac77-41a0-a642-16ad44da3855 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Stealing watermarks of large language models via mixed integer programming,
Reference 41
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 761aaf7b-3953-4a74-995b-dea8f5b81598 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Can watermarks survive translation? on the cross-lingual consistency of text watermark for large language models,
Reference 42
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation a9f7d773-9aae-4b4c-a6a9-a8e97877789a · outbound
Character-Level Perturbations Disrupt LLM Watermarks On Evaluating The Performance of Watermarked Machine-Generated Texts Under Adversarial Attacks
Reference 43
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation f3ff928d-3c0c-4544-bc3b-eb3e2a033914 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Silverspeak: Evading ai-generated text detectors using homoglyphs,
Reference 44
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 83e67408-48f4-47d0-9c06-5f3d00c9a010 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Watermarks in the sand: impossibility of strong watermarking for language models,
Reference 45
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 446f6627-c982-4311-a849-1648856c03e0 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Watermark under fire: A robustness evaluation of llm watermarking,
Reference 46
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 82c4f5a0-2b51-4e6b-835b-e0196476444f · outbound
Character-Level Perturbations Disrupt LLM Watermarks Sslguard: A watermarking scheme for self-supervised learning pre-trained encoders,
Reference 47
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7ba6361f-37c2-4ab9-861b-aa5a11034f7d · outbound
Character-Level Perturbations Disrupt LLM Watermarks Entangled watermarks as a defense against model extraction,
Reference 48
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation a518a146-2387-42d3-ad09-49313477f8c5 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Exploring the limits of transfer learning with a unified text-to-text transformer,
Reference 49
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2c7283d7-325b-49e1-ac60-945dc8da4060 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Synthid
Reference 50
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 7c481b3a-6f21-4141-8a4a-20099744aeca · outbound
Character-Level Perturbations Disrupt LLM Watermarks Terms of use,
Reference 51
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 888b75bd-7e67-4ff7-a710-539909486d53 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Optimizing adaptive attacks against content watermarks for language models,
Reference 52
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 532e6623-8cfb-4399-8b10-9d69ce010d5c · outbound
Character-Level Perturbations Disrupt LLM Watermarks Markllm: An open-source toolkit for llm watermarking,
Reference 53
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5ebde10f-8818-4b65-9628-7e3fc2812608 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Software Framework for Topic Modelling with Large Corpora,
Reference 54
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 8079ca2f-7778-4935-8656-362c6ccf748a · outbound
Character-Level Perturbations Disrupt LLM Watermarks AuthorMist: Evading AI Text Detectors with Reinforcement Learning
Reference 55
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 17d1b733-e8c0-4662-88c5-e027750bdcd4 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Bleu: a method for automatic evaluation of machine translation,
Reference 56
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ad71345a-99b2-4dda-a860-b192500809ba · outbound
Character-Level Perturbations Disrupt LLM Watermarks Rouge: A package for automatic evaluation of summaries,
Reference 57
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation c9135b88-33aa-44c1-9676-f03f0eb886ba · outbound
Character-Level Perturbations Disrupt LLM Watermarks The cefr levels
Reference 58
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2ded87d2-d40f-494a-a90e-7e672deb6a0b · outbound
Character-Level Perturbations Disrupt LLM Watermarks No free lunch in LLM watermarking: Trade-offs in watermarking design choices,
Reference 59
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 36368584-233b-4e00-9b85-40d1fc8c28aa · outbound
Character-Level Perturbations Disrupt LLM Watermarks New ai classifier for indicating ai-written tex,
Reference 60
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation ac015a89-3fbc-4f7b-8369-27e998cde758 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Ai detector by grammarly
Reference 61
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation d492b1c6-5759-4b8c-9aef-ca60198a7e40 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Gptzero’s ai detection technology
Reference 62
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation bd70620d-e1db-403e-8409-964d3466b614 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Bert: Pre-training of deep bidirectional transformers for language understanding,
Reference 63
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1852bfb4-a84b-401e-843f-68f7227f0ebf · outbound
Character-Level Perturbations Disrupt LLM Watermarks Languagetool
Reference 64
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 6635be85-c3ec-4b7c-b9f1-e6027b16a80e · outbound
Character-Level Perturbations Disrupt LLM Watermarks Python tesseract
Reference 65
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 94603794-456f-49da-ae57-f6817e1f2a8d · outbound
Character-Level Perturbations Disrupt LLM Watermarks Unicodedata
Reference 66
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 27693111-82a8-4098-bf69-763669810bde · outbound
Character-Level Perturbations Disrupt LLM Watermarks Results are reported forn= 5,10, and15, using both token- level and character-level perturbations across five watermark- ing schemes
Reference 67
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 11a52c92-b1d3-4a3b-8614-fd49f5be26b2 · outbound
Character-Level Perturbations Disrupt LLM Watermarks We compare both token-level (dashed lines) and character-level (solid lines) perturbations
Reference 68
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5b3f3546-5755-4d2d-8d3a-d65d2fbb1f25 · outbound
Character-Level Perturbations Disrupt LLM Watermarks As introduced TABLE XV: Performance of reference detectors with three levels of data augmentation across five watermark schemes
Reference 69
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 3a20e059-ac93-4eb6-958c-611393b97e47 · outbound
Character-Level Perturbations Disrupt LLM Watermarks Major Revision
Reference 70
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 1b166493-c74c-4c53-9781-266e1cd21bfc · outbound
Character-Level Perturbations Disrupt LLM Watermarks We recom- mend downloading the code using the following command: git clone https://github.com/plll4zzx/CharacterRemoval4WM
Reference 71
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.
Observation 61a40ad6-0521-4c50-9d3e-e998b30e21ea · outbound
Character-Level Perturbations Disrupt LLM Watermarks Unresolved cited work
Reference 72
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation be7ba321-042e-441d-8d9f-2a6d5dd1cb3d · outbound
Character-Level Perturbations Disrupt LLM Watermarks Unresolved cited work
Reference 73
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 602d3fd1-93ae-49ac-9d77-b6519a501039 · outbound
Character-Level Perturbations Disrupt LLM Watermarks real- newslike/*
Reference 74
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 5c90de35-de9d-4060-8c8b-694dcd33c20b · outbound
Character-Level Perturbations Disrupt LLM Watermarks /../dataset/c4/realnewslike
Reference 75
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 0dc88317-ae32-4354-b1ec-0980abe5b17e · outbound
Character-Level Perturbations Disrupt LLM Watermarks KGW" --num_epochs 15 \ --rand_char_rate 0.15 --rand_times 9 \ --llm_name
Reference 76
Source-reported events for the cited work
Unavailable: canonical work link unavailable.
Observation 2ab31816-21f1-4680-ae47-4be4ff3b9c28 · inbound
DP-NCB: Privacy Preserving Fair Bandits Character-Level Perturbations Disrupt LLM Watermarks
Reference 44
Source-reported events for the cited work
No event found in the named queried sources as of 2026-08-09T06:31:02.800959+00:00.